I’m intrigued today by the recent flurry of social media posts I’ve seen about AI-generated art. It’s that time, I guess. There was that big rush earlier, and then a quiet lull. Now folks are posting again. Since I follow a lot of writers and a few artists I see the flurry as hand-wringing about the use of AI art, and the idea that using it will result in artists starving. Or, more likely, if you use it, you are starving an artist!
I suppose that’s possible, but the more I get into it, the less I think that the destruction of artists is a likely outcome of the AI phenomena. Just as I think the advent of AI storytellers are unlikely to put “real people” writers out of business.
I should disclose up front that this new wave of conversation comes as I’m reviewing proofs of a project I’ve undertaken to pair science fictional haiku with AI-generated art. It’s a book of haiku that I wrote, and then fed into an AI art generator to see what it would come out with. The project has been both really interesting, and great fun. By that, however, you can take it that I’m actively working with said engines of destruction. So, paint me as one actively starving the artists. I think that’s wrong (for reasons I’ll get into), but I so get the angst. On the other hand, I will also disclose here that I am specifically not using the names of any specific artists in my prompts, for reasons that I think will be obvious to those who are paying attention. Maybe I’ll go into that at a later date. I’ve written about AI art before. I’m sure I’ll write about it again. But for purposes of this little entry, I figure this disclaimer should suffice.
To me, that bottom line is this: the AI artist (be they visual, written, or audio) is likely to wind up being simply another source of entertainment. Just like all the other sources. When AI starts writing fiction (which is essentially possible now but maybe six months or a year from being really robust), then it will step into the entertainment marketplace I work in and compete with James Patterson and Stephen King and Nora Roberts and … and me. They all have their readerships, as I have my tiny one.
Note that above I used the term “entertainment” rather than art. I may think I make art, and I may be right. But the general public generally does not care about the same kinds of things me and my writer/artist/musician friends do. They care mostly about being entertained or enthralled or whatever. Readers go to the brands/writers they love, and that’s pretty much that. So, will AI artists steal all the brands? That’s the doomsday concern I see in the conversation of the day. The answer is: Maybe, but I doubt it.
Readers are not stupid. They will buy something they like and not buy something they don’t. They tend to want to support writers. Same thing with artists. And musicians.
AI entering the market as competition will change everything, of course. But change is not destroy. Some writers/artists/musicians will get crunched. Others will thrive. Some of that will be luck, more of it will be the ability to adapt and persist. It’s possible the market for “real people” art will shrink, but it’s also possible it will grow. No one, and I mean now one, actually knows what is going to happen.
On the business side, as an Indie writer, I can say that (as long as I have the skill and time to fiddle with making my own covers) the current price of using AI generators is much cheaper than other choices. But there are several pieces of thought that need to be overlaid here, not the least being that I’d expect that, even without legal actions going on that will eventually see artists paid for work the AI engines are using to train themselves up, access to these engines will soon grow more expensive. That’s capitalism for you, right? The cost/benefit analysis for indie publishers suggests that AI art as it is today is the way to go, but my view is that it’s not going to be so slanted as it is right now for too much longer.
Beyond that, I’ll point to a survey I spent some time with this morning in which a company released a survey of a very large set of independent authors. Among the questions was one about whether writers used professional cover designers or not. The answer was: it depends on how much money they make. Well, duh, right?
In that line “as long as I have the skill and time to fiddle…”the time I have available is as highly relevant as the skills required to do the work. This survey showed clearly that when indie writers are struggling early in their careers, they already do most of their own cover work (just as they also do most of their own editing—or shop it out to friends who will do it for free). But, unless an independent writer just loves doing cover work, this same survey shows that once they get enough revenue to shop cover design to someone else, a large percentage do it. This is because the economics of the decision are obvious. These writers make more money writing new stuff than they can save by doing their own artwork.
The existence of AI generated art will alter that market (of course it will), but in the terms of the independent publisher (or traditional, too, I suppose) the first question will be “can I make more money by contracting cover design out?” When the answer is yes, the next question is: “do I hire a physical artist, or do I hire a prompt-maker.” If the “real person artist” does work that sells books (especially assuming the cost of AI generated art rises), they’ll keep working. If they don’t, then they won’t. Realize, too, an AI prompt is not a book cover. A book cover includes font work and some element of graphic design at a minimum. These are not rocket science, but they do require (today) a human in the loop.
I personally am right in the middle of that kind of business decision right now, so I feel these pressures pretty clearly. As an independent publisher, the core of the business decision is based on profit and loss. Personally, I tend to enjoy the fact that I do a lot of my own covers. I like that the product I put out is “all mine.” But my schedule says I’ll be publishing 15-20 books next year. So, as I succeed up, time demands I at least take a look elsewhere.
There’s a schedule consideration, too (and that leads some to get upset because an AI generator can work relatively quickly), but realize that at least in the publishing business, the schedule is built around the book’s publication date. To keep working a visual artist will have to (1) sell books with their covers, and (2) be able to hit writers’ schedules. I don’t write fast enough to need immediate turn around on my covers. It’s all good. As long as artists can make covers that sell books, and complete them in sync with a writer’s ability to create books, it will all work out. Cost, quality, and schedule, right?
This is also true of writers, for that matter. It’s always been true. If your stories don’t draw readers, if they aren’t priced to market, and if you can’t keep making those stories rapidly enough to make your readership happy, you’re going to struggle.
But Ron, you’ll feel differently when it’s stories that AI is taking over, right?
Well…I dunno. I do sympathize with the sense of loss various artists are feeling—especially though who see their work sucked up to “train” the AI. It’s a strange dynamic. I get it.
And when the time comes that AI comes fully into writing fiction, well, we’ll see what really happens, but I don’t see it making a difference in my own world, though. People who read my stories already have a million-million more options to go to. Many of them giving their work away for free. A million more isn’t going to make a noticeable difference. And to be sure, the number of people making a living in independent publishing over the years continues to grow exponentially (as far as I can see).
Yet, when I release a new book, some of those people drop other forms of entertainment to read my work.
Will AI incursion make it harder to make a living in any particular creative field? Maybe. Even probably. But as I look at the world around me, I just don’t see it being catastrophic.
If I were being political here, I’d say this is basically Real World Example #287 of how the world’s economy is shifting due to automation, and why we will eventually be driven to the concept of Universal Basic Income of some flavor. But that’s a story for another day.
I expect that the various artificial intelligence systems out there will find niches of their own. And some people will go to them often. You can argue, then, that the market for “real people art” will shrink by whatever degree that slice winds up being. I don’t see it being 100%, though. In the realistic reality of the near term I don’t see it being super large at all. Given the survey I noted above, the market for AI-generated art (outside the kinds of novelty projects I’m messing with where the medium is actually the message) will be mostly newer writers with a little computer graphics skills, or writers with very high graphics skills already, and who enjoy doing their own work—which is a market that is already not using professional cover artists at high rates.
The cottage industry of AI prompters will grow, of course. And while access to AI generators is cheap, they’ll make headway into space “real people” artists are filling. But when the cost catches up, the drive to go to “real people art” will be greater.
Where will the equilibrium wind up?
I have no idea. And neither do you.
We can agree that it’s going to change a lot of the creative economy, and that means artists and writers and musicians need to be focused and probably nimble. But right now, over the long haul, I’ve come to doubt it will be completely apocalyptic for artists of any particular kind.